Re: Implications and Issues in Converting UML to RDF/OWL When Co nstructing HL7 RIM Ontology

Hi Chimezie,
I just wanted support the implications and issues with HL7 standards that
you describe by sharing some experiences working with CDISC standards and
FDA database model for clinical data. And also linking to an intersting
upcoming workshop to develop a Clinical Trial Ontolog.

> As a messaging standard, HL7 RIM isn't expressive or comprehensive enough 
> for use as a basis for a Computer-based Patient Record for Clinical
Research.

We came to the same conclusion when exploring CDISC SDTM 1] submission
standard, which are (weakly) connected to HL7 RIM, and FDA JANUS model 2]
for a future clinical datawarehouse (to store data submitted uisng the SDTM
standard). 

> [HL7] best suited as a messaging interlingua between systems than as a 
> primary representation model.

We also identified inherent problems as CDISC's standards are focusing on
exchange of data per clinical study, and not on making data recombinant
cross clinical studies.

> The conclusion that what is needed is a seperate 'Reference Ontology of
> the Healthcare Domain' and a 'Model of Healthcare Information' is well 
> founded.  
> ... and suggest that what is needed is first a common (domain-independent)

> framework for a CPR (expressed jointly in OWL and with a uniform, 
> accompanying XML syntax) which captures all the aspects of longitudinal, 
> problem-oriented patient data that are common across  various domains.

The author of the critical HL7 RIM paper, Barry Smith, is also one of the
people behind the upcoming workshop to develop a Clinical Trial Ontolog


1] http://www.cdisc.org/models/sdtm/
2] http://www.fda.gov/oc/datacouncil/Janus.zip
   http://crix.nci.nih.gov/projects/janus/
   http://gforge.nci.nih.gov/projects/janus
3] http://bioontology.org/wiki/index.php/Workshop_on_Clinical_Trial_Ontology


	Kerstin Forsberg
> Senior Information Architect
> Information Strategy, Clinical Information Science
> AstraZeneca
> 
> 
>

Received on Friday, 1 September 2006 18:16:43 UTC